Early hands-on: OpenAI's Astra excels at browser use and 3D, faster and cheaper than Fable
bindureddy · x · 2026-09-04
Developer Bindu Reddy shares an early readout of OpenAI's unreleased Astra model:
- Strong at 3D modeling tasks
- Browser use is its superpower, with impressive long-running agent loops
- Faster than Fable, with fewer iterations needed, making it cheaper to run
She calls it a very exciting model — but worries fear-mongering could push OpenAI to nerf its capabilities before launch.
Related event: OpenAI launches flagship GPT-6 Astra to rave reviews and pricing debates(190 posts)→
More from Models
- Opus 5.5 tops OpenRouter spending in 20 of 29 task categories — mmurph · 2026-10-21
- Meta team builds open-source 30B LLM from scratch in months, hails 'second era' — TimDarcet · 2026-10-21
- DFlash drafter for Muse Glimmer-30B: 5 layers, blockwise 16-token spec decoding — TimDarcet · 2026-10-21
- EvoPress quantization shrinks 30B model from 56GB to 19GB at near-bf16 accuracy — TimDarcet · 2026-10-21
- Radar chart: Muse Glimmer-30B benchmarked against Gemma4-31B and Qwen3.6-27B — TimDarcet · 2026-10-21
- Privacy evals: Muse Glimmer-30B leaks less than Qwen3.6-27B on CIMemories — TimDarcet · 2026-10-21